Talking about adherence

Kane et al are to be commended for advocating an individualized approach to medication adherence. They note that “clinicians generally spend too little time assessing and addressing adherence attitudes and behaviours”. That clinicians devote little time to this in routine clinical practice is significant, as the quality of the clinician-patient relationship is known to influence adherence to treatment. A recent meta-analysis found the odds of a patient adhering to treatment to be 2.16 times greater if there is a good doctor-patient relationship (1). This association has also been found in mental health care (2) and specifically in the treatment of psychosis (3). However, to date, interventions to improve adherence have tended not to focus on the prescribing clinician-patient relationship.

The clinician-patient relationship is reflected and negotiated in clinician-patient communication. A central component is shared decision making, which is receiving increased attention in medicine. Its importance is well established in the medical literature, with a small but growing evidence base in the mental health field. In the treatment of schizophrenia, shared decision making has been found to help patients feel more informed about their illness and treatment, improve satisfaction with care (4) and reduce hospitalization (5). However, there are few observational studies of shared decision making in psychosis.

In video recorded outpatient psychiatric visits, there was, on average, one medication decision in every visit, lasting about two minutes (6). Hence, this is a central part of routine reviews, where both the clinician and the patient can negotiate, influence each other's views and come to a decision that is more or less likely to be followed by the patient once he/she leaves the room. That these decisions take only two minutes reflects how little time is spent on this central aspect of care. Patients in these follow-up visits, half of whom had a diagnosis of schizophrenia, were not greatly involved in the decision making process, with a mean score of 12.5% (i.e., 6 out of a total possible score of 48). Although this appears to be higher in first visits (7,8), there is wide variation between psychiatrists in the extent to which they involve their patients in decisions about medication. Further research is warranted to identify what influences such wide variation and how shared decision making can be enhanced in practice.

The challenges to and benefits of improving adherence are well documented. It is also worth pausing to consider the benefits of non-adherence from the patient's perspective. Non-adherence is not always an irrational decision, with evidence suggesting that some patients do quite well without maintenance medication. Increasingly, many people are uneasy about being told that they will be taking antipsychotic medication for the rest of their lives. With an increasing consumer movement, people wish to take more responsibility for their health. They raise concerns about the ongoing unpleasant side effects of medication, how they interfere with their ability to fulfil key social roles and the risks of long-term antipsychotic use to their physical health.

If patients do wish to discuss reducing or discontinuing medication, this can be problematic. In a study of patients coming off antipsychotic medication, 38% of them were not comfortable disclosing this to their doctor and came off medication without telling them (9). More patients were unwilling to disclose that they intended to come off antipsychotics than patients coming off antidepressants. This is a riskier scenario than patients sharing this information and staying in touch with services so their progress can be reviewed. This highlights the importance of joint discussion about the benefits and risks of adherence and non-adherence with each individual and ensuing negotiation about a way forward. Moreover, this discussion needs to be ongoing, as an individual's mental health and personal circumstances vary over time. Depending on the culture of services, this discussion may be more or less difficult for clinicians. There is considerable institutional pressure on psychiatrists to adopt a cautious approach and real dilemmas in facilitating trial periods without medication. For some, it is too risky.

In a study of communication and adherence in the treatment of schizophrenia, patient participation in the form of asking questions and requesting clarification of the psychiatrist's talk was associated with better adherence to medication six months later (10). As with shared decision making, there was considerable variation between psychiatrists in how often their patients requested clarification. Hence, evidence suggests that there is good and poor communicative practice, which impacts on adherence. However, this needs further unpacking, so that specific communication skills can be targeted in training and peer supervision.

Different medications and dosages have different effects on individuals. This is reflected in psychiatrist-patient discussion of the patient's subjective experience of current and past medications, to inform changes in type and dose. Kane et al point to the potential of new technologies, e.g., a digital feedback system recording when medication is taken, along with physiological measures, to directly assess adherence and also act as interventions to enhance adherence. Such technologies also offer other exciting opportunities to use this information to tailor medication type, doses and frequency to an individual patient in order to identify the most tolerable and therapeutic regimen. Given the adverse side effects of antipsychotics, this would be a welcome advance.

Many factors influence adherence. Many of these factors are impossible to intervene in and change. Clinician-patient communication can be observed, and it is possible to intervene to change communication. However, the focus should be on adherence to joint decisions rather than adherence to medication per se.